You are on page 1of 14

Scheduling for Success with Critical Chain

by Rob Newbold, September 2010

Abstract
The Critical Chain scheduling approach described in this paper can be used to create “good” schedules that have provided substantial benefits in speed, predictability, and efficiency across many industries. 1 The first section of this paper explains the importance of a good schedule for managing an individual project, and describes what is needed for a schedule to be “good.” The second section, “Constructing the Schedule,” shows how ProChain builds Critical Chain schedules with the needed attributes. The third section, “Using the Schedule,” describes how good schedules can be used to gain significant benefits.

1. A Good Project Schedule
A project schedule is only one part of a successful project management methodology, but a good schedule can provide the basic framework for excellent project management. 2 Some of the benefits that accrue from good schedules include: • • • • • Credible schedules create consensus about priorities, enabling teams and organizations to reduce multitasking and to align on where to focus. Up-to-date information on where to apply attention and resources anticipates problems and keeps them from spiraling out of control. Better handoffs and better preparation for critical tasks create faster, more efficient flow of work. Portfolio decisions improve when you have realistic commitment dates and a better understanding of resource requirements. Improved predictability, for individual projects and across the portfolio, allows better decisions by both internal decision makers and investors.

In other words, a good schedule is essential to synchronizing project work in ways that help the organization. Before considering how to construct and use a good project schedule, we must first ask what that means, because not just any schedule will do. A good schedule must have a few key attributes. First, it must be credible to all those associated with the project. Otherwise it will not be used. Second, building that credibility will require that the schedule and the processes that produced it be owned by project team members and management. If they don’t own it, it won’t be part of their solution. And third, the schedule must be used. That means, for example:

Numerous examples are available; see, for example, (Newbold 2008, p.5). Note that this paper focuses on the scheduling of individual projects. In many situations, the addition of multi-project scheduling can produce significant additional benefits. See also (Newbold 2008, chapter 13). 2 The importance of project planning in general is gaining more and more formal recognition, as shown by its move to prominence in (PMI 2008). Copyright © 2010 by ProChain Solutions, Inc. All rights reserved. 1

1

3 Used to communicate status – status of tasks. The network should be built to achieve well-defined. and portfolios. and use the schedule to make informed tradeoffs. get what you got. the more there will be confusion about priorities. Inc. company-level objectives. “Drawings Rev. Without the ability to make valid predictions.1 Step 1: Build the Network In order to build a credible. we will plan for “business as usual”. For example. projects. It 3 4 For a detailed description of the logic behind this. preferably with verbs in the names and supporting detail attached. This section describes such a process. we shouldn’t expect things to change. Inability to set priorities through credible schedules results in increased emphasis on fixed due dates and increased multitasking. be real and well-defined.” Maybe everyone knows what Rev. suppose you have a task called. rather than the behaviors we see. All rights reserved. created. Most projects have a very small number of objectives that we really care about – usually one. or purchased? What are the drawings of? With such an ambiguous name. and how they link together. consisting of five steps: • • • • • Step 1: Build the Network Step 2: Level the Load Step 3: Calculate the Critical Chain Step 4: Protect Against Uncertainty (The Project Buffer) Step 5: Synchronize the Non-Critical Chain Tasks (Gating) 2. 2 . When possible and sensible. Used in analyzing possible actions and making meaningful predictions. see (Newbold 2008. If we plan for the behaviors that are already in place.” Copyright © 2010 by ProChain Solutions. we start by building a dependency network: a model of the required tasks and resources. management is flying blind. start with your best estimate of what should happen. Depending on the complexity of the project and the time available to schedule it. tasks should have resources assigned. Constructing the Schedule In order to construct a good Critical Chain schedule we need a systematic and rigorous process. great detail may not be feasible. broadly owned project schedule. pp. Tasks in a schedule should be real and actionable.Scheduling for Success with Critical Chain • • • Used to set priorities. 2. The network should be based on the collective input of the project team. the more commitments a team makes. both within and between projects. 1. What are you delivering to the clients? What is making you money? Pick your objectives and your project endpoints carefully. resulting in conflicts and multitasking. When you’re not sure. adapted. 1 Drawings are. 4 Therefore the project schedule should embody the behaviors we want to move towards. but what’s the work? Are the drawings supposed to be delivered. or who has to do it? The connections between project tasks should. how could you determine how long it will take. Or: “Do what you did. whenever possible. 14-15).

3 Step 3: Calculate the Critical Chain The next step in the Critical Chain scheduling process is to identify the set of tasks that determines the project’s earliest feasible end date. Safety will be added in Step 4. Figure 1 shows a simple project plan with three tasks (A. However. These tasks are called the “Critical Chain. Otherwise. Task A is required to finish before Task B can be started. and Tasks B and C must both finish before the endpoint can be achieved. In addition. such as vendor deliveries. there may be parallel sets of tasks that are all critical. This process of resolving resource contention is called “resource leveling” or “load leveling. the arrows indicate links between tasks. Many software packages. Each box represents a task. to protect the entire project. including ProChain’s. we must resolve any contention that tasks might have for scarce resources. 3 . with no safety time added. which means using “focused durations” or “touch times” for task duration estimates. Other modeling elements can be added as necessary. those high-confidence durations will be used to adjust the contributions of individual tasks to buffers.Scheduling for Success with Critical Chain should assume no multitasking. 2.” During the scheduling process. Note that leveling only improves credibility when focused durations are used.2 Step 2: Level the Load Because we’re looking for credibility. Tasks B and C both need the resource “Artist. the critical path would not represent the most important tasks to focus on and the timing of the endpoint would not be credible.” Task A Task B: Needs resource “Artist” Task C: Needs resource “Artist” Figure 1: Basic Project Plan Tasks A and B comprise the traditional critical path. task restrictions such as “start no earlier than…” may be used to indicate information that can’t easily be modeled as tasks. we may significantly affect the schedule’s credibility and utility. if we assume there are no limitations on available resources. and C) feeding into a milestone (the black diamond). In a network with many parallel chains of tasks. All rights reserved. After leveling on the Copyright © 2010 by ProChain Solutions. provide optimization functions that can help to minimize the pushing that can result from resource leveling. if we have only one Artist resource. as described below. For example.” During the leveling process we also consider limitations on the availability of resources or timing of tasks. the Critical Chain must take into account resource limitations. Inc. In the figure.” They usually represent the most important places to focus management attention. B. nonfocused durations often include time to work on multiple things. High-confidence durations may be used to show how long tasks might take as a “worst case” or “90% probability. 2. In order to maintain credibility. Each task’s duration should reflect the amount of work expected to be done on it by whoever is doing the work.

especially if we have truly used focused durations. The Critical Chain consists of Tasks B and C (and.” 5 If we could only announce one date publicly and wanted to have a safe commitment date. We need to calculate a range of times over which the project’s deliveries are likely to be made. All rights reserved. when we make it explicit. the endpoint). Task A Task C: Needs resource “Artist” Task B: Needs resource “Artist” Figure 2: Leveled Project Plan and Critical Chain 2. second is a phenomenon we call “integration risk. we need to decide on a credible duration for it.Scheduling for Success with Critical Chain Artist resource. because we haven’t created a credible schedule. we have the picture shown in Figure 2. Copyright © 2010 by ProChain Solutions. we’d probably announce the “latest finish” for the project. the “earliest feasible” delivery may be highly unlikely. but because announcing it would jeopardize the credibility of the entire schedule. we break its duration into two components. This range of times. In order for the Project Buffer concept to be useful. Figure 3 shows the earliest feasible finish of a small project. Unfortunately. It’s called a buffer because. from “earliest finish” to “latest finish. To do that. our resource-leveled schedule shows where the endpoint will fall in time. this could be called the “earliest feasible” delivery. of course.” The Project Buffer is a time range. not just because the “earliest finish” has a low probability of coming true.” These components are described in the following sections. we have created and validated a project network and identified the most important “Critical Chain” tasks. because in most projects there is significant variation in how long things take. we can use it to protect project endpoints from variations in the time required to attain those endpoints. It also shows an estimated “latest finish” time. Inc. as determined by the Critical Chain (the red tasks). Representing the delivery as single point in time is almost certain to be wrong. 5 4 . First is variability due to tasks on the Critical Chain.4 Step 4: Protect Against Uncertainty (The Project Buffer) By performing steps 1 to 3. Project Buffer Project Start Earliest Finish Latest Finish Figure 3: The Project Buffer If we wanted to be in big trouble. That means we haven’t finished scheduling. we’d publicly commit to the “earliest finish. Let’s start by considering components of that range. This would put us in trouble. In addition.” is also known as a “Project Buffer.

All rights reserved. 2. Copyright © 2010 by ProChain Solutions. suppose that in Figure 2. If the high-confidence duration isn’t specified. and avoids the kind of under-estimation that we’ve commonly seen with (for example) Monte Carlo simulation. 6 The variability component is typically calculated as a percentage of the Critical Chain’s duration. For example. Sometimes tasks along the Critical Chain have more or less variability than average and this needs to be reflected in the Project Buffer. 5 6 . for example.2 Integration Risk If we looked only at the Critical Chain tasks. we would leave out important sources of variation: where the non-Critical Chain tasks integrate with the Critical Chain.4. This gives a simple and predictable model that experience has shown to be effective. As a simple example. If instead Task C had been modeled with a high-confidence duration of 40 days. So it can be useful to specify high-confidence or worst-case durations. the time to grow a particular bacteria culture may have little variability.g. If a Critical Chain task has a high-confidence duration specified. rather than 10 days. some software development projects).4. This can be a simplification.Scheduling for Success with Critical Chain 2. in order to indicate the unusual variability of such tasks. so when using 50% buffer sizing the Project Buffer duration would be 20 x 50% = 10 days. it’s defaulted to be twice the focused duration. Figure 4. some construction projects) and larger percentages can work with projects having higher variability (e. The Critical Chain length is 20 days. there are times when it may make sense for a high-variability task that is not on the Critical Chain to contribute to the variability portion of the buffer.1 Critical Chain Variability Some of variability in the endpoint will come from tasks on the Critical Chain. smaller percentages can work with projects having lower variability (e. we can calculate a safe completion date (“latest finish”) for the project. the programming of a new software module may have a great deal. Inc.g. Task C’s variability contribution would be (40 – 10) = 30 days. Consider. That is. in addition to focused durations. While we normally use 50% of the Critical Chain to size project buffers. the difference between high-confidence and focused durations is used in sizing the Project Buffer. and where there are multiple Critical Chain tasks that integrate together. Task B and Task C both have tenday durations. if we analyze the possible variation on Critical Chain tasks. The resulting Project Buffer size would then be (30 + 10) x 50% = 20 days.

5 x . If the probability of completing Tasks A and B by the dotted line is . Tasks A and B each have a 50% chance of finishing after one day or one toss. must both complete before work can start on Task C. Task C may still be delayed.” Copyright © 2010 by ProChain Solutions. Task A takes a little more time than expected and Task B a little less time). Multiplying 70% by 70% for Tasks A and B would give a 49% (roughly 50%) chance starting of Task C at that point. 8 To calculate a precise time at which Task C would have a 50/50 chance of starting. Suppose each toss represents a day of work.25.5 Task A Task C Task B 2. We call this potential delay “integration risk.5 = . 6 . All rights reserved.5 x . Assume the probability of finishing A or B here = . because both Tasks A and B have to get heads in one toss.5 (50%). This is shown in Figure 5.125.5 = . and the task is completed when you get heads. the probability of starting Task C at the early time would be . it can’t start until the later of the two predecessors completes. which have identical durations. then the probability of starting Task C is the product of Tasks A and B’s probabilities: .25 Figure 4: The Effect of Integration Risk Suppose Tasks A and B. The chance of starting Task C after one day is 50% x 50% = 25%. Then the probability of finishing both A and B here = probability of starting C here = . More tasks being integrated would require more allowance for integration risk.Scheduling for Success with Critical Chain 1.” because it represents a real risk of delay at an integration point. The time between the two dotted lines is additional buffer time we need to allow for the integration risk.5 x . 7 This integration phenomenon implies that even if the durations of Tasks A and B are on average correct (say.5 x .5 = . assuming precision is possible. and Tasks A and B are independent. Inc. If we had three such tasks being integrated. 8 This phenomenon is also known in project management literature as “merge bias. we’d need to find the point where each predecessor has about a 70% chance of finishing. and it can be managed as a risk. 7 It’s easy to simulate the durations of Tasks A and B using coin tosses.

the recommended alternative. the integration risk protection is not incorporated in the Project Buffer itself. There are various mechanisms in common use to calculate integration risk.Scheduling for Success with Critical Chain 1. 7 . Critical Chain has traditionally used a concept called the “Feeding Buffer” to decouple non-Critical Chain tasks from the Critical Chain. Then the probability of finishing both A and B here = probability of starting C here = . it can be used to identify the key integration points and to analyze the highest-variability chains that feed them.7 Task A Task C Task B 2.7 ≅ . They need to use their experience. All rights reserved. Some people use Monte Carlo simulation. some people allow the Feeding Buffers to push out the Critical Chain. their schedule. Copyright © 2010 by ProChain Solutions. Both components are important in protecting the project’s commitment dates from disruptions. “gating” them. 9 By breaking the Project Buffer’s duration into these two components and analyzing them. and analysis tools to answer questions like. we gain a certain amount of visibility into the reasons for its size. The simplest way of doing that is by scheduling them as soon as possible. with more consistent answers and without running multiple trials. “Are the Project Buffers adequate to protect against likely problems and delays?” and “Does the overall schedule reflect the variability that we expect in our projects?” 9 When using Feeding Buffers to calculate integration risk. is discussed below as part of Step 5. 2.4. The ProChain mechanism is both intuitive and easy to analyze. In that case.3 The Full Project Buffer The project buffer is a single entity (usually represented in a schedule as a task) containing both a variability component and an integration risk component. but in gaps in the Critical Chain. ProChain provides a mechanism that calculates results that are similar to Monte Carlo simulation. We have found this approach to be adequate and easy to explain for simple projects. although we’ve found that this approach also creates complexities in understanding and analysis. Inc. Ultimately – as with the rest of the project schedule – the Project Buffer durations should be evaluated for credibility by the project team and management. It can also produce illogical results and be difficult to analyze in complex.5 Integration Risk Figure 5: Calculation of Integration Risk Note that non-critical chain tasks should be scheduled early enough that their disruptive effect on the Critical Chain is minimized. Assume the probability of finishing A or B here = . real-world projects.7 x .

We’ll look at schedule analysis first. They should be synchronized with the Critical Chain in such a way that we minimize the needed Project Buffer. However. under the assumption that work that’s already “in process” should be finished with a minimal amount of multitasking.Scheduling for Success with Critical Chain 2. This idea of starting tasks “when needed” (rather than as early or late as possible) is called task gating. The integration risk calculations tell us how late we can schedule non-Critical Chain tasks while still minimizing integration risk. A project manager usually has the authority to start work on gated tasks earlier if he or she deems it prudent. is generally represented by a changing percentage complete of the “buffer task. Project Buffers are typically represented as tasks. 10 Task A is scheduled early enough to avoid integration risk (and therefore an impact on project completion). However. Inc. 10 8 . Gating can help delay project expenses until they really need to be incurred. and analyzing the likely impact of various actions. Tasks are typically not gated if work has already started. communicating status. non-Critical Chain tasks do have slack time. by definition. and consequently we have some latitude in scheduling them. It also helps to minimize the number of active tasks. they can’t logically be scheduled any earlier. but we haven’t discussed the other pieces: setting priorities. and hence temptations to multitask or micro-manage. but otherwise it’s as late as possible. we don’t need to compensate for possible delays at the integration point by adding to the Project Buffer. All rights reserved. Gating time Time needed to eliminate Integration Risk Task A Task C: Needs resource “Artist” Task B: Needs resource “Artist” Project Buffer Figure 6: Full (Gated) Critical Chain Schedule 3. However.” Copyright © 2010 by ProChain Solutions. with a Project Buffer at the end represented as a task. and there can be advantages to scheduling them later.5 Step 5: Synchronize the Non-Critical Chain Tasks (Gating) Critical Chain tasks are scheduled as early as possible. Figure 6 shows the full schedule created from Figure 2. noncritical tasks may not be needed right away. which occurs when delays cause tasks to push into the buffer they feed. Consumption of Project Buffers. If a chain of non-critical tasks can be moved early enough that there is no integration risk where it feeds into the Critical Chain. Using the Schedule At this point in the process we have taken important steps towards building a credible schedule (one of our “good schedule” requirements). because a project team typically In software scheduling tools. This “early enough” time is known as the gating time. they aren’t normal tasks in that they have no work and their scheduled times don’t change when the schedule is updated. The obvious solution is to schedule all tasks as soon as possible.

chances are you’ll have to analyze several to noticeably reduce your integration risk. 3. 9 . However. Besides correcting errors. This analysis is not usually a good way to spend much of our time.1. We can use the Critical Chain to determine where our validation efforts will be most fruitful. These analyses can provide great value in suggesting ways to speed up project completion. because integration risk is by definition based on multiple chains. Copyright © 2010 by ProChain Solutions. For example. which means we also bring in our commitment date.3 Analyze Project Buffers Do we need all of the Project Buffer that has been allocated? In other words. All rights reserved. we should still validate that the integration risk is credible and that it’s not significantly over. 3. resources. By looking at the integration points.2 Analyze Critical Chain Reducing the duration of Critical Chain tasks will reduce the duration of the entire project. do we understand and accept where the uncertainty in our commitment date is coming from? Do we need more time? The variability component of the Project Buffer is normally determined by the duration of the Critical Chain tasks and (if specified) their high-confidence durations. These elements can be analyzed to determine whether the variability component seems reasonable.Scheduling for Success with Critical Chain does significant analysis to improve schedule quality and timing before communicating priorities or status outside the team.1. and re-scheduling allow us to analyze tradeoffs and make good decisions.1. because in-depth analysis efforts to reduce the duration of an invalid network will usually waste time.1 Validate Is the network a valid representation of what we expect to happen? Does the Critical Chain make sense? Are the links. We conduct this analysis early.or under-stated. this analysis should include real challenges to the schedule assumptions and the status quo. and durations reasonably correct? Fix the obvious problems before doing anything else. 3.1 Schedule Analysis There are a number of areas we can analyze in order to understand and minimize the time required to complete the project. analyzing. 3. where you must integrate five parts. Inc. we can also explore the impact of integration risk on the start and finish dates of Project Buffers. because shortening the Critical Chain will usually result in shortening the Project Buffer and moving it to the left. This helps reduce the time spent scrutinizing less critical portions of the network. Can we bring in additional resources? Are all the planned features essential to project delivery? Can we do work “at risk” and thereby break links? What are the tradeoffs? The results of scheduling.

Copyright © 2010 by ProChain Solutions. Updates to remaining durations of the tasks themselves. Re-calculate the Critical Chain. Apply any needed updates to the schedule. Based on the updated schedule data. and/or The current (“as-of” or “status”) date. we need to compute the current picture: how much Project Buffer has been consumed. All rights reserved. When even a long a project is coming down to the wire. A one-year project might be updated weekly. In Figure 8. which tasks are now most critical (the current Critical Chain). although in some environments it’s helpful to update a schedule more often. daily standup meetings can be useful. 10 11 . or calendar time. The solid red and black bars in Figure 8 show how the task schedules have changed from the originals. the horizontal black line in Task C indicates that some work has been done. Task A Task B Project Buffer Task C Status Date Figure 7: Updated Schedule Some of the changes that occur after a schedule update are illustrated in Figure 7. This includes resolving resource contention based on the current picture and re-calculating needed gating times. calculate the amount of incursion into the buffers by the endpoints that feed them. in this example. but it would be reasonable to update a two-month project daily. there are typically three types of changes that can affect the schedule: • • • Adjustments to the structure of the work needed. Re-calculate task times based on current information. and possible changes to task gating. In Figure 7. • • A simple rule of thumb would be to update schedules at least every 2% of the overall project duration. • • • Move the current or “status” date out to its new time. links.Scheduling for Success with Critical Chain 3. We find that it’s usually adequate to do this on a weekly basis. 11 When an update occurs. Based on the re-calculated times. it still consists of Tasks B and C. resources.2 Creating and Communicating Project Status In order to maintain ongoing credibility. a schedule and its associated buffer status must be kept up-to-date. Inc. but no less than weekly. This is converted to buffer % consumed. such as added or deleted tasks. which is based on the schedule from Figure 6. the solid red bars represent the Critical Chain.

For each schedule update. chapter 10. and green regions are slanted. with the left. tasks scheduled to start tomorrow will almost always be lower priority than tasks scheduled to start today. All rights reserved. shown in Figure 8.Scheduling for Success with Critical Chain Figure 7 shows some Project Buffer consumption due to the delay of Tasks C and B. including how the tasks impact project endpoints. yellow. The boundaries between the red. stable priorities help people to focus attention where it’s needed and reduce multitasking. Since we gate tasks so that they don’t start until they’re needed. Which do you work on next? There are many factors that can be considered in prioritizing tasks. 11 . The variety of factors and the breadth of 12 For more information on this. a new point is added to the picture. A popular means of communicating project status is the Fever Chart. Inc. the yellow zone indicates potential trouble. respectively. Now suppose you have a number of tasks that could start today. Figure 8 has four data points. representing the initial schedule and three updates. and the green zone indicates that things are probably on track. including current task status and task gating. When prioritizing tasks.and right-hand lines representing 0% and 100% complete. the relative priority of the tasks’ projects. % Buffer Consumed 100 50 0 0 50 100 % Project Complete Figure 8: Fever Chart 3. 12 The horizontal dimension is the percentage of the Critical Chain that has been completed. and the current status of those projects. This reflects the fact that late in a project.3 Setting Task Priorities Task prioritization is an essential output of any good scheduling process. with the top line representing no buffer left (100% consumed). the first place to look is at the current schedules. If the project’s status point falls in or above the red zone. The vertical dimension is the percentage of Project Buffer that has been consumed. the project is likely in trouble. and helps you to extrapolate the potential for the project to be late. In that way the Fever Chart provides a good sense of project progress to this point. Clear. Copyright © 2010 by ProChain Solutions. see (Newbold 2008). less buffer is needed to maintain confidence in the safe commitment date.

a critical chain task feeding it would have a PTI of 20%. See. This scheduling and updating process produces what we defined at the beginning of this paper as a “good” project schedule: • • • • The Critical Chain scheduling process is designed around ownership and credibility. we’ve found “Percent Task Impact” (PTI) to be adequate. During both network creation and updating. It is defined as the impact a task would have on the Project Buffer if the task were to be delayed. All rights reserved. time must be included in the Project Buffer to protect against integration risk. non-critical chain tasks would have lower PTI values. identify the chain of tasks that determines project completion. and relative task criticality. Project status is communicated through Project Buffer consumption and the Fever Charts. you must build a credible project network. If non-Critical Chain tasks can’t be moved sufficiently early.” and protect the customer (delivery) against disruptions along the Critical Chain by creating a Project Buffer. Task priorities. task status is communicated through measures of how close a task is to being on the Critical Chain. schedule the non-critical tasks earlier to protect the Critical Chain itself against merge bias or “integration risk. helps to ensure that Critical Chain projects meet their schedule.” For individual projects. predictability and efficiency. Summary In order to construct a good Critical Chain schedule. a project buffer is 20% consumed. and cost objectives. from the team-based network building to project buffering to regular project updates. With the credible schedule and up-to-date project status. 5). we can analyze possible management actions and predict their impact. the schedule should be analyzed for opportunities to complete the project more quickly. Inc. p. This process. (Newbold 2008. thus focusing our time and resources in areas where they’ll do the most good. All these factors help companies that adopt the Critical Chain scheduling process to gain significant improvements in speed. buffer status. for example. scope. 4. within and between projects. for example. coupled with ongoing buffer analysis and timely management actions. 12 13 . Copyright © 2010 by ProChain Solutions.Scheduling for Success with Critical Chain possible considerations suggest that we can’t possibly find a single numeric measure that can be used to prioritize tasks “optimally. If. 13 There are many documented results of the effectiveness of Critical Chain scheduling. are set using information about tasks’ impacts on the Project Buffers. level the load on resources. It is a combination of available slack and existing buffer consumption. The Critical Chain updating process keeps the picture up-to-date by calculating updated task times.

you have a much better chance that people will respect the need for focused work. divided by the buffer duration. A “focused duration” for a task is the time expected to be spent on that task. and durations. if it were worked start-to-finish with an absolute minimum of interruptions. The dimensions are Project Buffer % consumed versus % Project Complete – essentially. Inc. people in your environment may not be able to focus on their work effectively. If you plan for people to spend focused time by specifying focused durations in your schedules.Scheduling for Success with Critical Chain Key Terms Buffer consumption Buffer consumption is a percentage calculated during a schedule update. It protects your customer from variations in the time needed to perform the work. 100 – (current Critical Chain length) / (original Critical Chain length). The project buffer end dates are usually published as project commitment dates. resources. 13 . Integration risk Work at an integration point may only start when all the components being integrated are present. High-confidence (low-risk) durations These are the “worst case” or “90% probability” durations for tasks. thereby delaying the start of work at the integration point. it should be on the Critical Chain tasks. Integration risk (also known as “merge bias”) is the risk that one or more components feeding an integration point will be late. Credible dependency network This forms the basis of a Critical Chain schedule: an activity-based model of the required work with validated links. Focused durations Today. Project Buffer The Project Buffer is the range of time over which you expect a project delivery to be completed. If you’re going to focus anywhere. All rights reserved. times 100. Fever Chart The Fever Chart is a picture of how project status has changed over time. It is the amount of time that an endpoint’s optimistic finish goes past the start of the Project Buffer it feeds. The difference between the low-risk and focused durations tells us the contribution of individual tasks to the buffers they feed. Copyright © 2010 by ProChain Solutions. Critical Chain The Critical Chain is the set of tasks that determines the timing of a project’s endpoint(s). Each point on the Fever Chart represents a schedule update.

A Guide to the Project Management Body of Knowledge. All rights reserved. The Billion Dollar Solution: Secrets of ProChain Project Management.” if delaying them is unlikely to affect the project completion date. 2008. Lake Ridge. Newtown Square. VA: ProChain Press. 14 . References Newbold. Inc.Scheduling for Success with Critical Chain Task gating “Gating tasks” are tasks that have not been started and that have no predecessors. Robert C. 2008. Copyright © 2010 by ProChain Solutions. PA: Project Management Institute. Fourth Edition. Project Management Institute (PMI). These tasks will be scheduled to start later than “as soon as possible.